AI Biotech Rising Star Ensem Therapeutics Files IPO Following $1.33B Deal and Pipeline Breakthroughs

AlphaFold2, as a groundbreaking AI tool, has a profound impact on the development of life sciences. Most intuitively, several large-scale financings and transactions related to AI drug discovery have emerged, and new AI drug discovery companies have also been born under this background, with Ascend Medicine being one of them.

In November 2023, Ensem Therapeutics emerged as a dark horse in the AI pharmaceutical industry. BeOne Medicines announced a collaboration with Ensem Therapeutics to obtain the global exclusive license for the latter's differentiated CDK2 inhibitor, ETX-197. Ensem Therapeutics will receive payments of up to $1.33 billion, including upfront and subsequent milestone payments.

This deal acted like a deep-water bomb, triggering industry attention towards this previously low-profile AI pharmaceutical company and also sparking a new round of discussions about the AI pharmaceutical sector.Following the CDK2 inhibitor deal, Ensem Therapeutics, the rising star in AI-driven drug discovery, continues to make strides. Its second potential "best-in-class" allosteric inhibitor targeting PI3Kα mutant proteins is nearing the candidate compound stage, with the third candidate compound soon to follow.

The establishment of Onsure Pharmaceuticals benefited from CBC Group's forward-looking judgment of the industry and the rapid formation of a senior founding team.

The birth of AlphaFold2 allowed Fu Wei, CEO of CBC Group, to see the development potential of AI technology in the pharmaceutical field, and he hoped to incubate an AI-driven pharmaceutical company. In the AI pharmaceutical sector, CBC Group first recognized the potential of AI in researching protein dynamic pockets and targeting difficult-to-drug sites.

Dynamic pockets of proteins can undergo deformation or adjustment based on the functional requirements of the protein and its interactions with ligands. In the field of drug design, understanding the dynamic behavior of these pockets can help scientists design drug molecules that bind more effectively and modulate protein activity.

However, due to the complexity of protein structures, identifying dynamic pockets in proteins is a significant challenge, creating a formidable barrier for the druggability of many targets. Chen Liang, Vice President of healthcare investment at CBC Group, mentioned:The greatest value of a new technology is its ability to solve problems that traditional methods could not address.If AI technology can be leveraged to identify dynamic pockets of proteins and design drugs for difficult-to-target sites, it will represent an important direction for AI to break through traditional methods and truly empower new drug development.

Based on this concept, Dr. Cao Wuxiong, Operating Partner of CBC Group, invited Dr. Jin Shengfang to co-found Antelope Biopharma in Boston. At the time, Dr. Jin happened to have entrepreneurial aspirations, which aligned perfectly with CBC Group's vision of incubating AI-driven pharmaceutical companies.

In 2021, Ensem Therapeutics was founded, bringing together numerous seasoned pharmaceutical and AI technology scientists, which facilitated the development of its AI-powered technology platform, Kinetic Ensemble.^®The Rapid Birth.

Dr. Shengfang Jin, the co-founder, serves as the President and Chief Executive Officer of Ascend Pharmaceuticals. Prior to this, she worked at the renowned gene-editing company Editas Medicine. Earlier in her career, Dr. Jin joined Agios Pharmaceuticals as the third employee, where she was part of the founding team that led the development of Agios’ portfolio of small-molecule drugs for cancer metabolism and rare genetic diseases, contributing to the advancement of IDHIFA.^®、TIBSOVO^®、PYRUKYND^®The launch of three first-in-class new drug products.

In this process, Agios has established a strong collaboration with Schrödinger, a pioneer in CADD, and the three drugs led by Dr. Jin in research and development are also supported by Schrödinger's technology.

Jeff Kutok, Ph.D., serves as the Chief Scientific Officer of Angsheng. With over 25 years of leadership experience in biotechnology fields such as clinical practice and translational scientific research, he has held the position of Chief Scientific Officer at renowned biopharmaceutical companies like Epizyme and Infinity Pharmaceuticals, driving the development progress of multiple small-molecule drugs.

In addition to the two leading scientists, Jin Shengfang and Jeff Kutok, Ensem Therapeutics continues to strengthen its team capabilities in new drug development and AI technology iteration: Vice President of Medicinal Chemistry Liu Tao, DMPK Vice President Raj Nagaraja, and CMC Vice President Eric Simone are all key members of the former Agios team. Their joining has solidified Ensem Therapeutics' new drug R&D capabilities. The AI/ML and computational chemistry teams are co-led by former Director of Computational Chemistry at Agios, former Head of AI/ML Computational Chemistry at EISAI, Hao Minghong, and former Senior AI/ML Scientist at Atomwise, Hossam M. Ashtawy. Mengdi Wang, a tenured professor at Princeton University and a senior visiting scientist at Google DeepMind, serves as a core AI advisor providing guidance.

OncoMed Pharmaceuticals is developing with a strong momentum of accumulation and burst.

After more than two years of low-key technology and experience accumulation, at the end of 2023, Ensem Therapeutics reached a pipeline collaboration with BeOne Medicines for its CDK2 inhibitor ETX-197, breaking the record for the largest amount in AI pharmaceutical-related transactions in Asia with a deal worth $1.33 billion.

The excellent background of Ensem Therapeutics' founding team was one aspect that caught BeOne Medicines' attention, while the other aspect was the solid strength of the CDK2 inhibitor ETX-197.

According to Dr. Lin Ying, Vice President of Ensem Therapeutics, after discovering the dynamic binding pocket of CDK2 through AI technology, Ensem successfully designed ETX-197 with unprecedented speed. After undergoing various testing stages, including in vitro assays, animal DMPK experiments, and toxicology studies, the key data for ETX-197 has been remarkable. Compared to other companies' CDK2 candidate compounds or clinical compounds, ETX-197 demonstrates excellent performance in preclinical safety, activity, and therapeutic window.

The communication between Ensem Therapeutics and BeOne Medicines began in May 2023, when BeOne Medicines conducted a comprehensive evaluation of ETX-197. Despite being a newcomer, Ensem Therapeutics still won the favor of BeOne Medicines with its solid data and strength.

After the partnership was reached, Dr. Wang Lai, Senior Vice President of BeOne Medicines and Head of Global R&D, stated: "BeOne Medicines is committed to developing novel molecules to improve the treatment landscape for cancer patients. There is a significant and unmet clinical need in the field of breast cancer, and this collaboration aligns perfectly with our strategic focus in this area. Ensem Therapeutics' CDK2 inhibitor complements our internally developed CDK4 inhibitor, which is currently in Phase 1 clinical trials, with the potential to enhance the therapeutic effects of existing CDK4/6 inhibitors in specific breast cancer patients. This collaboration also strengthens our early-stage R&D pipeline in breast cancer and other solid tumors.”

This transaction cooperation also achieved the validation of Angsheng Pharmaceutical's own technology platform and product pipeline.Next, Ensem Therapeutics will welcome its second candidate compound in the second half of this year, an allosteric inhibitor targeting the PI3Kα mutant protein.Currently, orthosteric inhibitors (such as alpelisib and inavolisib) targeting this site on the market lack sufficient selectivity for wild-type PI3Kα due to their binding site being located in the ATP pocket, leading to severe side effects in clinical settings, such as hyperglycemia. Therefore, it is particularly crucial to develop highly specific and low-toxicity inhibitor drugs.

Onward Secures Another Swift Success with Kinetic Ensemble^®The platform identified a new metastable conformation and designed candidate compounds targeting this allosteric pocket.This compound demonstrates stronger cellular activity and preclinical in vivo efficacy compared to other allosteric inhibitors in early clinical development or at the clinical trial application stage, as well as higher selectivity for wild-type PI3Kα and other subtype proteins (PI3Kγ, PI3Kβ, and PI3Kδ). Additionally, compared to these competitors, Ensem Therapeutics' allosteric inhibitor exhibits enhanced activity in degrading mutant p110α protein in animal models/in vivo pharmacology experiments. All these preclinical characteristics are likely to confer superior activity and safety to this compound, along with significant potential to become a "best-in-class" drug globally.

ETX-197 is Ensem Therapeutics' first product through Kinetic Ensemble^®The technology platform identifies dynamic binding pockets and ultimately designs compounds. This undoubtedly demonstrates the development potential of Ensem Therapeutics' AI technology platform.

Currently, Ensem Therapeutics is focusing on building and optimizing two major AI models: EnsemNet.^®With EnsemGen^®Mainly used for precise prediction of protein dynamic pockets and generation of ligand small molecules that bind to them. Meanwhile, this technology platform is driving the development of multiple new drugs by Ensem Therapeutics.

EnsemNet^®It is a pre-trained deep learning model mainly used to predict hidden ligand-binding pockets in proteins and distinguish between matching and non-matching compounds. The model is based on a deep learning approach using geometric representations of large protein models and small molecules, allowing for transfer learning across numerous dynamic protein pockets to discover entirely new ones. The binding pocket of ETX-197 was identified using this method. Currently, Ensem Therapeutics' team is also working on EnsemNet.^®Continuous technical upgrades will be carried out, and its 2.0 version will fine-tune parameters to more accurately predict the hidden binding pockets of proteins.

EnsemGen^®It is an AI technology that generates protein dynamic conformations and complex structures based on the Boltzmann Generator and more advanced, flexible diffusion models (Diffusion Model). Since proteins are dynamic structures, they can have many different conformations. Drug discovery requires identifying biologically meaningful conformations that ensure stability and capturing ligand compounds associated with these conformations. After efficiently and rapidly discovering pockets of practical significance, Ensem Therapeutics can use these pockets for drug design, virtual screening, and further drug development.

OncoImmune has been continuously iterating and upgrading its two major technology platforms.While developing its own computing platform, the Angsheng AI/ML team aligns with cutting-edge technological breakthroughs in the field.For example, the recently launched AlphaFold3 (AF3) represents a significant leap forward in the construction technology of protein-biomolecular complexes. AF3 introduces several notable improvements, such as optimizing the primary deep neural network inherited from its predecessor, AlphaFold2, focusing on extracting the most critical information required for accurately assembling protein-biomolecular complexes. While these technological advancements are encouraging, it is important to note that AF3 is not yet fully open-source. It tends to predict static structures similar to those in the Protein Data Bank (PDB), rather than capturing the dynamic behavior of biomolecular systems in solution. AF3 still requires further refinement to enhance its accuracy in predicting specific protein-ligand complexes to truly aid practical drug discovery efforts. Although AF3's Nature paper demonstrates predicted protein-ligand complexes with commendable accuracy, Ensem Therapeutics can also generate similar models using a combination of existing methods. In summary, AlphaFold3 marks a major technological advancement in the field, and the Ensem Therapeutics team will fully leverage AF3’s auxiliary potential. As the platform gradually matures, Ensem Therapeutics' AI technology will be applicable to the drug development of various small-molecule modulators, offering limitless potential for pipeline expansion.

Currently, with the completion of the ETX-197 transaction partnership and the upcoming candidate compound targeting PI3Kα mutant protein in the second half of this year, the third candidate compound in Ensem Therapeutics' pipeline will soon follow. Numerous biological research projects in the exploratory phase are currently undergoing early testing. In the future, Ensem Therapeutics will continue to produce "best-in-class" or "first-in-class" small molecule candidate compounds annually. On the path of AI-driven drug discovery, Ensem Therapeutics shows great promise, and we look forward to their next breakthrough achievement.